1278 THE BELL SYSTEM TECHNICAL JOURNAL, NOVEMBER 1954 



time required to reach the critical velocity is, from efjuation (1), 



/ coo/o 118.5(7.36) nnrir J 



' = G7 = 13,500 = °°''* ''"""^' 



experimental speed-time curves — normal torque 



Experimental velocity versus time curves were obtained for the fly-bar 

 and drive-bar governors constructed to the specifications listed in Tables 

 I and III. Data used in determining the true speed versus time picture 

 for the experimental governors was taken from photographic traces ex- 

 posed on a recording oscillograph. A thin disc having 36 radial slots 

 spaced every 10° was fastened to the end of the governor shaft. Light 

 detected through the slots of the moving disc by the element of a photo 

 tube was used to deflect one of the strings of the oscillograph. The trace 



Table III — Refer to Figs. 5 and 7 



of this string appeared on the photographic paper as a distorted sine 

 wave. The distance between two successive wave peaks represented 10° 

 of rotation of the governor. By noting on the trace the time between 

 peaks, it was possible to determine the average velocity of the governor 

 at 10° intervals after release of the finger- wheel, or start of rotation of the 

 governor mechanism. The experimental speed curves for the drive-bar 

 and fly-bar governors are plotted on Fig. 6 along with the theoretical 

 speed curves. 



It was assumed in the theoretical analysis that while accelerating up 

 to the critical velocity, the governor assembly rotates as a simple fly- 

 wheel. This requires that the velocity increase linearly. The theoretical 

 and experimental velocity curves for both type governors during the 

 initial accelerating period show the fly-wheel assumption to be justified. 

 The slope of the velocity curve, or rate of acceleration is generally con- 

 stant. 



For that portion of the theoretical and experimental curves which show 

 speeds from the critical velocity to 98 per cent of rated speed, agreement 

 is not too clearly defined. The theoretical curve is naturally smooth in 

 shape. An oscillating type characteristic appears in the experimental 



